Part One: Semantic confusion: how using the vocabulary of human cognition to describe transformer-based language models obscures our understanding of artificial intelligence and the human mind.